Abstract :

Every Day, Millions Of Clothing Reviews Are Posted On E-commerce Platforms, Reflecting Customer Experiences And Preferences, And Nearly 80% Of Online Shoppers Rely On These Reviews Before Making Purchase Decisions. This Rapid Growth Of User-generated Content Has Created A Strong Demand For Intelligent Systems That Can Automatically Predict Product Ratings And Analyze Textual Reviews To Support Customer Satisfaction, Inventory Optimization, And Marketing Strategies. However, Traditional Machine Learning Models Rely Heavily On Manual Feature Extraction And Struggle With Noisy Data, Overfitting, And Limited Labeled Datasets. To Overcome These Challenges, This Work Proposes A Multi-model Framework For Simultaneous Rating Prediction And Recommendation Analysis In The Clothing Industry. The Proposed Algorithms Include Multi-Task Neural Network (MTNN) With Extra Trees Classification And Regression Trees (CART), Restricted Boltzmann Machine (RBM) With CART, Extreme Gradient Boosting (XGBoost) With CART, And Gradient Boosting, All Applied To Both Regression For Rating Prediction And Classification For Recommendation Tasks. The MTNN Architecture Uses Shared Feature Representations With Task-specific Branches, Enabling Efficient Joint Learning Across Tasks, While Ensemble CART Models Improve Robustness And Stability. Regularization Techniques Are Employed To Enhance Generalization Performance. Experimental Results Show That The MTNN With Extra Trees CART Achieves The Most Accurate Predictions, Providing Reliable Insights Into Customer Sentiment, Product Quality, And Overall Satisfaction In The Clothing E-commerce Domain.

Published:

24-4-2026

Issue:

Vol. 26 No. 4 (2026)

Page Nos:

3045-3058

Section:

Articles

License:

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

How to Cite